Skip to contents

Find the shortest abbrevation to retain unique values

Source: R/jam_abbrevations.R
shortest_unique_abbreviation.Rd

Find the shortest abbrevation to retain unique values

Usage

shortest_unique_abbreviation(
  x,
  retain_contig_numbers = TRUE,
  verbose = FALSE,
  ...
)

Arguments

x

character vector

retain_contig_numbers

logical, default TRUE, whether numbers at the end of an abbreviated string should remain contiguous.

  • When TRUE, the goal is not to split a numeric value in the middle of the number.

  • When FALSE the string will be abbreviated at the first position of uniqueness.

...

additional arguments are ignored.

Value

character vector named using unique values in x, and whose values are the shortest abbreviated substrings which maintain consistent uniqueness.

Details

This function is intended to abbreviate factor levels used in statistical contrasts to the smallest substring that uniquely represents the unique entries provided in x.

For example, c("one", "two", "three", "four") would be converted to c("on", "tw", "th", "fo").

The default retain_contig_numbers=TRUE will attempt to retain numeric values at the end of a string, to avoid splitting the number at an intermediate position. This option only applies when the character substring is not already unique before encountering the numeric substring.

* For this input:

c("a", "p6", "p12", "p21") the output keeps the contiguous numbers together: c("a", "p6", "p12", "p21")

  • For this input: c("a", "b6", "c12", "d21") only the first character is retained, because it is already unique: c("a", "b", "c", "d")

Todo

  • Consider some method to retain contiguous numbers at the end of a long string, while abbreviating the long string.

    • For this input: c("adult", "prenatal6", "prenatal12", "prenatal21") the ideal output would be: c("a", "p6", "p12", "p21")

    • To be fair, I do not know how to describe this logic. It may required breaking into words by character/non-character breakpoints, then applying substring to each?

See also

Other jamses utilities: choose_annotation_colnames(), combine_sestats(), contrast2comp_dev(), fold_to_log2fold(), intercalate(), list2im_opt(), list2im_value_internal(), list_to_sestats(), log2fold_to_fold(), make_block_arrow_polygon(), mark_stat_hits(), matrix_normalize(), merge_statdf_all_test(), point_handedness(), point_slope_intercept(), shrinkDataFrame(), shrink_df(), shrink_matrix(), sort_samples(), strsplitOrdered(), sub_split_vector(), update_function_params(), update_list_elements()

Examples

x <- c("a", "p6", "p12", "p21");
shortest_unique_abbreviation(x)
#>     a    p6   p12   p21 
#>   "a"  "p6" "p12" "p21" 

shortest_unique_abbreviation(x, retain_contig_numbers=TRUE)
#>     a    p6   p12   p21 
#>   "a"  "p6" "p12" "p21" 

x1 <- c("male", "female");
shortest_unique_abbreviation(x1)
#>   male female 
#>    "m"    "f" 

x2 <- c("Control", "Nicotine");
shortest_unique_abbreviation(x2)
#>  Control Nicotine 
#>      "C"      "N" 

x3 <- c("Control", "Nicotine10", "Nicotine12", "Nicotine20");
shortest_unique_abbreviation(x3)
#>      Control   Nicotine10   Nicotine12   Nicotine20 
#>    "Control" "Nicotine10" "Nicotine12" "Nicotine20" 

x4 <- c("one", "two", "three", "four");
shortest_unique_abbreviation(x4)
#>   one   two three  four 
#>  "on"  "tw"  "th"  "fo"